Research on Web Log Analysis System
Zhao Hai-ting · Computer Technology and Development · 2011
Web log analysis system can not only improve the Web site structure and improve Web server performance,but also identify the user's preferences,satisfaction,identify potential customers and enhance the competitiveness of Web services.The stages of Web log mining are described,and a Web log analysis system is designed and implemented.The shortcomings of traditional frequent itemsets mining algorithm and sequential pattern mining algorithm are analyzed.According to the characteristics of log data,the user attributes are added into the generation process of frequent item sets,effectively reducing the number of candidate items.According to the characteristics of the candidate set,by round of compressed database.ApiroriAll continuous sequence introduced into the algorithm and the process of candidate set.An improved algorithm is implemented.In the experiment,the efficiency of improved algorithm and traditional algorithm is compared,the effectiveness of the improved algorithm is proved.